Human Perceptions of Robotics in Agriculture

In recent years, the agricultural sector has been witnessing a technological change characterized by the automation of various activities also due to the use of some robotic technologies. As for the field of application, the pioneer- ing solutions available on the market aim at performing non-selective oper- ations. A key point of success in this field of research is the implementation of cognitive human processes in the algorithms driving robot actions. Howev- er, the perception that people have of the use of robots in agriculture is still little studied. The current research is a preliminary study aiming at describ- ing human perception of different robotic solutions that are already availa- ble on the market and/or that will be potentially developed in future. An on- line survey analyzed different case studies encompassing several robotic plat- forms that are expected to perform selective vs. non-selective and high risk vs. low-risk operations in viticulture. The study evaluated human perception of 49 students mainly enrolled at the faculties of education and psychology with a moderate knowledge in agriculture. The results showed that people prefer the use of humans over robots to perform selective agricultural activities, thus possibly reflecting greater reliability associated to humans in making decisions and performing selective tasks. On the other hand, when comparing the dif- ferent robots, the humanoid robot ‘replaces’ the human, i.e., people consider it as the robot to be used in selective activities requiring more complex deci- sions. This is also confirmed when analyzing the perception on the effect that the introduction of humanoid robots may have in different domains of agri- culture: there is a preference for the use of humanoid robots for performing selective tasks in both the safety and quality domains. Finally, autonomous ve- hicles are preferred to humanoid robot in terms of increased work productivi- ty and reduced risk. Generally, these findings allow us to hypothesize that peo- ple place human and robots on two different ontological levels with respect to their mental characteristics, as reported by the participants’ preference for us- ing robots in tasks that do not require complex decisions. This consideration is further supported by the few differences found between humans and robots in both high and low risk non-selective tasks.